BACKGROUND: A system to rapidly assess influenza vaccine effectiveness (VE) is needed. The aims of this research are to 1) generate annual estimates of direct VE for prevention of laboratory-confirmed influenza in a noninstitutionalized population-based cohort; 2) provide rapid assessment of VE at regular intervals during the influenza season; and 3) provide estimates of vaccine effectiveness in higher risk age groups. The Marshfield Clinic (MC) infrastructure offers three key strengths to conduct this research: l) a well-defined, stable population cohort of about 50,000 people living in 14 zip codes (Marshfield Epidemiologic Study Area, or MESA); 2) a comprehensive, real-time, internet-based immunization registry that includes all public and private immunization providers serving this population; and 3) a functioning syndromic surveillance system to rapidly identify incident cases of medically-attended acute respiratory illness (MAARI). METHODS: This will be a prospective cohort study using a fixed cohort based on the MESA population. The cohort will be divided into the following age strata: 6--23 months, 24 months through 8 years, 9--17 years, 18 49 years, 50--64 years, and 65+ years. The cohort will be defined on November 1, 2004, and influenza immunization status will be monitored in real-time. Each age strata will be further stratified into high-risk and normal-risk based on the presence or absence of certain chronic diseases. We will identify and enroll cohort members who have MA.ARI during the influenza epidemic period. We will attempt to enroll all MAARI cases in the priority age groups for whom influenza vaccine is recommended: children 6 months through 23 months old, and adults > 50 years old. We will employ a sampling strategy for MAARI occurring in other age groups (2 years--49 years). A nasal or nasopharyngeal swab will be collected from enrolled patients for viral culture. Patients with MAARI will be identified and enrolled using two procedures: 1) same-day recruitment during a clinic visit to Pediatrics or Urgent Care, and 2) next-day enrollment using syndromic surveillance, which will identify patients with MAAR/visits on the previous day at all MC locations. ANALYSIS: Efficacy of the influenza vaccine will be estimated as VE=I-RR where RR denotes the relative risk of laboratory-confirmed influenza in the vaccinated group compared to the unvaccinated group. Initially, all members of the influenza study cohort will be classified in strata defined by MAARI status, vaccination status, age group and high- risk status. In each MAAR/stratum, enrolled subjects will be assigned an analytic weight equal to the number of people in the study cohort for the stratum divided by the number of enrolled subjects in the stratum. Overall VE estimates will be adjusted for age group, high-risk status and resource utilization. We will compute cumulative VE estimates at weeks 4, 8 and 12 of the influenza season. In addition, separate 4-week estimates for weeks 5-8 and 9-12 will be computed. We will use a variety of analytic procedures to identify potential sources of bias and confounding. Interim and annual estimates of VE with 95% confidence intervals will be provided to CDC, including overall VE and VE for each of the priority age groups.